Factors associated with dropout from treatment: An exploratory study.

AIM: To evaluate the factors associated with treatment dropout among patients attending the psychiatric outpatient services.
MATERIALS AND METHODS: Seventy-two patients who dropped out from treatment were contacted and they were compared with 200 regular attendees for sociodemographic and clinical variables, medication adherence, treatment satisfaction, attitude toward medication, insight, and therapeutic alliance.
RESULTS: Compared to "regular attendees," those who dropped out from treatment were significantly older, were more likely to be married, had higher age of onset, had longer duration of illness, received less supervision for medication at home, higher proportion of them continued to remain symptomatic, had more negative attitude toward medications, had poorer insight, were poorly complaint with medication, were less satisfied with the treatment provided, and had poor quality of therapeutic alliance.
CONCLUSION: This study suggests that dropout from treatment can be avoided by addressing issues of negative attitude toward medications, improving satisfaction with the treatment contact and enhancing therapeutic alliance. Copyright:

INTRODUCTION

Although many studies have focused their attention on the issue of nonadherence to psychiatric medications, research on nonadherence to clinical appointments and factors associated with dropout is much less extensive. The estimated dropout rates with outpatient psychiatric services vary considerably, ranging from 14% to 64% in different studies.[1] On an average a quarter to half of people, who miss an outpatient appointment completely disengage from the mental health services.[2] In the largest study of full disengagement, 31% of community psychiatric patients were lost to follow-up over the course of 1 year.[3] Over the course of 2 years, the documented attrition rates for unscheduled dropout have been reported to be as high as 50%.[4]

Factors which influence nonadherence to appointments or dropout from treatment can be broadly understood as environmental and demographic factors, patient-related factors, memory/cognitive problems, information and health beliefs, illness factors, and clinician and referrer factors.[56789101112131415] The most consistent demographic correlates of nonattendance/dropout across several studies have been younger age, male gender, impaired functioning, and different indicators of a socially disadvantaged status (unmarried, poorly educated, unemployed or low income patients, and from minority communities and low socioeconomic strata). Other factors such as the ability to pay for treatment, distance from the treating facility, lack of transportation, homelessness, substance use, severity of the illness, and living alone have also emerged as the correlates of nonattendance in certain other studies.[16171819] Environmental factors associated with nonattendance include the type of treatment-setting, longer delay in receiving the initial appointment, and longer intervals between appointments.[1520] Some patients cite unhappiness with treatment as their reason for nonattendance. The most common patient-related factor for nonattendance at follow-up appointments is forgetting the appointment, followed by being too psychiatrically unwell.[21] The proportion of psychiatric patients accidentally (unintentionally) missing an appointment appears to be higher than that reported for other medical specialties.[5] Other patient-related factors of nonattendance include high trait anxiety and lower social desirability scores.[22] Memory impairment may play a role in nonattendance as well as in subsequent engagement with treatment and services.[23]

Clinical correlates of missed appointments or dropouts are somewhat less certain. Almost every type of diagnosis including psychosis, mood disorders, neurotic disorders, and personality disorders has been linked with nonattendance in different studies.[2425] However, substance abuse or dependence when present is almost always associated with nonattendance.[26] There appears to be a bimodal relationship between severity of illness and attendance, with both low and high illness severity predicting nonattendance.[27] The duration of previous contact with psychiatric services also emerges as a consistent correlate of nonattendance.[28] Interestingly, if the diagnosis is unclear (or cannot be established), patients are more likely to drop out of care.[9] In addition, the dual-diagnosis group have a dropout rate, which is almost three times higher than that of the single-diagnosis group.[6]

In terms of clinician-related factors associated with nonadherence, quality of service provided and rapport between patient and clinician are important determinants. Studies suggest that poor communication between the referring practitioner and the patient may increase nonattendance at an initial appointment.[21] There is also an association between the quality of the referral letter and the likelihood of attendance.[2930]

Perhaps the most important interpersonal variable in predicting attendance is the quality of the therapeutic alliance and, more specifically, the degree of 'helpfulness' of the health professional as perceived by the patient.[31] In some studies, clinician-rated alliance is a stronger predictor of dropout than patient-rated therapeutic alliance.[3233]

The extent of nonattendance in psychiatry and its impact on the patient may be significantly greater than in most other medical specialties.[534] The consequences of missed appointment are more serious for people with severe and enduring mental health problems such as schizophrenia, schizoaffective disorder, and bipolar affective disorder than those with common mental disorders.[126]

There are few studies from India which have focused on the dropout rates of patients attending the psychiatric outpatient services of the general hospital psychiatric units,[353637383940414243] outpatient's services of mental hospital (NIMHANS),[44] rural psychiatric clinic,[45] and private psychiatric clinic.[46] Although there are differences in the definition used to define “dropouts,” the dropout rates reported in these studies have varied from 21.3% to 59.3%.[35363738394041424344454647] Another common finding from all these studies is that majority of the patients drop out after their initial visit. Some of these studies have also evaluated the factors associated with dropout from treatment, but there are very few common factors reported across all these studies.[363739404447] However, none of these studies have used any standardized instrument to assess variables such as treatment satisfaction, medication adherence, and severity of illness. In this background, this study aimed to evaluate the factors associated with treatment dropout among patients attending the psychiatric outpatient services.

MATERIALS AND METHODS

This cross-sectional study was conducted in a tertiary care multispecialty hospital in North India. Ethical clearance was obtained from the institutional ethics committees. All the participants aged ≥18 years were recruited after obtaining written informed consent. Participants aged ≥15 to <18 years provided written assent, in addition to the parental consent. The study sample comprised two groups, i.e., Group I (regular attendees) and Group II (dropouts).

To be included in the study, the participants were required to be aged ≥15 years and diagnosed with one of the mental disorders. Patients with an exclusive diagnosis of substance abuse or dependence, intellectual disability, significant cognitive impairment due to organic brain syndromes, not diagnosed with any psychiatric disorder, who did not complete all assessments, those who had no access to a phone, and who had registered but were not seen a psychiatrist at least once were excluded. A convenient sampling method was used.

At our setting, all the new patients attending the walk-in clinic are initially evaluated either by a senior resident or a faculty member and the diagnosis is made as per the International Diseases of Classification, tenth revision criteria.

The regular attendee group (N = 200) included participants who had continued to follow-up in the psychiatric outpatient for at least 6 months following their initial registration and attended at least 75% of their scheduled appointments. The dropout group comprised patients who had registered with the psychiatric services at the same time frame as the “regular attendees” and attended the psychiatric services at least once for treatment, but had not returned to the clinic within 6 month of the last visit. The hospital records (walk-in proforma/workup file/inpatient file) of these patients were reviewed to make such a conclusion. In addition, data were also extracted from the registration counter to cross-check the data. All the patients considered to be 'dropped out' were contacted telephonically and they were explained about the study. They were invited to attend the psychiatric outpatient to participate in the study. Those who agreed and attended the psychiatric outpatient were explained about the study again in detail and written informed consent was obtained prior to recruitment. A total of 200 patients were contacted, out of which 115 (57.5%) agreed to come for participation. However, only 72 (36%) came for participation and were recruited. Patients in the dropout group were asked to provide information, relevant to the time frame of their contact with our services.

The study participants were evaluated as follows.

Brief Adherence Rating Scale

This was used to evaluate medication adherence in quantitative terms. It provides a sensitive, reliable, and valid measure of antipsychotic/medication adherence in patients as compared to electronic monitoring.[48] Furthermore, this scale is simple, quick, and easy to administer making it applicable for clinical settings for assessing adherence to any kind of medications.

Satisfaction with treatment

Satisfaction with treatment among both patients and relatives was assessed using a 4-item rating scale devised by Shipley et al.[49] This was used as a proxy measure for the staff–patient relationship.

Drug Attitude Inventory-10 item version

The Drug Attitude Inventory-10 item version (DAI-10)[50] is the most commonly used and better validated self-report measure comprising ten “yes/no” statements reflecting patients' experiences, attitudes, and beliefs about medications.

Self-report attitude scale (SRAQ)

Attitudes toward medications were assessed using a self-report questionnaire (in Hindi).[51] It is based on an earlier instrument devised by Helbling et al.[52] It has 18 items rated on a three-point scale and examines attitudes to psychotropic drugs in general.

Scale to Assess Unawareness of Mental Disorder (SUMD)

The Scale to Assess Unawareness of Mental Disorder utilizes a semi-structured interview to rate discrete and global aspects of insight. It evaluates the three most relevant aspects of insight, viz. awareness of mental illness, awareness of the need for treatment and awareness of the social consequences of mental disorder. It has satisfactory convergent and criterion validities and can be used reliably with minimum training.[53] Separate ratings were done for the current level of insight and insight at the beginning of the treatment.

Scale to assess the therapeutic relationship

Scale to assess the therapeutic relationship (STAR) has been specifically developed to assess the relationship between multidisciplinary clinicians and patients with severe mental illness in community care settings. It has good test–retest reliability and internal consistency. The patient version has 3 subscales, viz., positive collaboration, positive clinician inputs, and nonsupportive clinician inputs. Completing the scale usually takes 5 min or less. Scores can be obtained for the total scale and subscales.[54] For this study, STAR-client version, which has 12 items, was used.

Global assessment of functioning scale (GAF)

The assessor rated the current level of functioning using the Global Assessment of Functioning scale (GAF).[55] The GAF, which forms Axis-V of the multiaxial DSM-IV, is used to global functional level on a 100-point single item scale.

Clinical Global Impression-Severity of Illness Scale

The assessor rated the current severity of illness using the Clinical Global Impression-Severity (CGI-S).[56] The CGI-S is an observer-rated scale that measures current illness severity on a 7-point scale, with the severity ranging of from 1 (normal) through to 7 (amongst the most severely ill patients).

Reasons for dropout

This was evaluated using an open-ended question, in which the patients were asked to give one or more reason(s) for dropping out from treatment.

Descriptive analysis was carried out using mean and standard deviation with range for continuous variables. Frequency and percentages were calculated for discontinuous variables. Comparisons were carried out using the Chi-square tests, Student's t-tests, and Mann–Whitney U-tests. The association between different variables was studied using Spearman's or Pearson's correlations.

RESULTS

Socio demographic profile of the dropout and regular attendee groups

Demographic profile of both the groups is given in Table 1. The mean age of the “dropout” group was significantly higher than the mean age of the “regular attendee” group. Compared to the “regular attendee” group, significantly higher percentages of patients in the “dropout” group were married [Table 1].

Table 1

Comparison of sociodemographic profile of the dropout group and regular attendee group

Variables	Frequency (%)/mean (SD)		χ2/t (P)
	Dropout group (n=72)	Regular attendee group (n=200)
Age (years)	40.23 (12.6)	35.5 (12.3)	2.75 (0.006)**
Sex
Male	37 (51.4)	90 (45)	0.868 (0.351)
Female	35 (48.6)	110 (55)
Marital status
Currently single	12 (16.7)	68 (34)	7.661 (0.006)**
Married	60 (83.3)	132 (66)
Education (years)	10.4 (3.78)	10.7 (4.5)	0.488 (0.626)
Education of patient
Up to high school	39 (54.2)	97 (48.5)	0.680 (0.410)
More than high school	33 (45.8)	103 (51.5)
Occupation of the patient
On paid employment	29 (40.3)	61 (30.5)	2.286 (0.131)
Not on paid employment	43 (59.7)	139 (69.5)
Income of patient
Up to 7322	50 (69.4)	145 (72.5)	0.244 (0.622)
>7322	22 (30.6)	55 (27.5)
Socioeconomic classa
Lower/middle	40 (55.6)	113 (56.5)	0.019 (0.890)
Upper	32 (44.4)	87 (43.5)
Religion
Hindu	46 (63.9)	123 (61.5)	0.128 (0.720)
Non-Hindu	26 (36.1)	77 (38.5)
Type of family
Nuclear	50 (69.4)	115 (57.5)	3.165 (0.075)
Nonnuclear	22 (30.6)	85 (42.5)
Locality
Urban	42 (58.3)	98 (49)	1.846 (0.174)
Rural	30 (41.7)	102 (51)
Distance from the hospital (km)
Up to 80	43 (59.7)	123 (61.5)	0.070 (0.791)
>80	29 (40.3)	77 (38.5)

aSocioeconomic class was defined as per the Kuppuswamy definition. SD – Standard deviation. **P<0.01

Clinical profile of the Dropout and Regular Attendee Groups

In both the groups, the single largest diagnostic group comprised the “other disorders,” which included all the disorders other than the psychotic and affective disorders [Table 2]. However, compared to the “regular attendee” group, significantly higher proportion of patients in the “dropout” group had “other disorders” and correspondingly lower prevalence of the patients with “psychotic disorders.” Those in the “dropout” group had significantly lower age of onset of the illness, longer duration of mental illness, and higher number of hospitalizations compared to those in the “regular attendee” group [Table 2].

Table 2

Comparison of clinical profile of the dropout group and regular attendee group

Variables	Frequency (%)/mean (SD)		χ2/t (P)
	Dropout group (n=72)	Regular attendee group (n=200)
Age at onset of illness	35.38 (12.80)	31.56 (12.05)	2.261 (0.025)*
Duration of illness (months)	55.84 (58.98)	49.78 (75.01)	5844# (0.018)*
Family history of psychiatric illness
Yes	24 (33.3)	44 (22)	3.627 (0.057)
No	48 (66.7)	156 (78)
Past psychiatric treatmenta
Yes	12 (16.7)	43 (21.5)	0.767 (0.381)
No	60 (83.3)	157 (78.5)
Primary diagnosis
Psychotic disorders	4 (5.6)	33 (16.5)	8.255 (0.016)*
Affective disorders	23 (31.9)	76 (38.0)
Other disorders	45 (62.5)	91 (45.5)
Comorbid physical illnessb
Present	13 (18)	30 (15.0)	0.371 (0.542)
Absent	59 (82)	170 (85.0)
Patient ever hospitalized in a psychiatric setup
Yes	3 (4.2)	2 (1.0)	FE=0.118
No	69 (95.8)	198 (99.0)
Total number of hospitalizations	0.04 (0.20)	0.01 (0.14)	6937.5# (0.03)*

*P<0.5; #Mann-Whitney U-value; aPast psychiatric treatment means seeking treatment from a psychiatrist; bin the dropout group, 5 patients had hypertension, 3 had diabetes mellitus, 1 had hypothyroidism, and 4 had other comorbid chronic physical illnesses; in the regular attendee group, 7 patients had hypertension, 4 had diabetes mellitus, and 15 had other comorbid chronic physical illnesses and 4 had hypothyroidism. FE – Fisher’s exact value; SD – Standard deviation

Treatment details of the Dropout and Regular Attendee Groups

When both the groups were compared, medications were being supervised in significantly higher percentage of patients in the regular follow-up group. Similarly significantly higher percentages of patients in the regular follow-up group were advised relaxation exercises and behavioral therapy, complied with the advice of relaxation exercises and dietary advice [Table 3]. When the number of visits were analyzed, it was seen that more than one third (37%; N = 27) in the dropout group dropped out from treatment after initial evaluation [Table 3].

Table 3

Comparison of treatment details of the dropout group and regular attendee group

Variables	Frequency (%)/mean (SD)		χ2/t (P)
	Dropout group (n=72)	Regular attendee group (n=200)
	Pharmacological treatment
Number of patient on nonpsychotropic drugs
Yes	19 (26.5)	35 (17.5)	2.62 (0.10)
No	53 (73.5)	165 (82.5)
Number of psychotropic medications prescribed	2.05 (4.45)	1.65 (0.69)	7100# (0.85)
Number of nonpsychotropic medications prescribed	0.38 (1.13)	0.23 (0.76)	6562# (0.11)
Total number of medications	3.48 (4.66)	1.88 (1.06)	6869# (0.53)
Dosing schedule of psychotropic medications
OD	52 (82.2)	161 (80.5)	2.13 (0.14)
BD	20 (27.8)	39 (19.5)
Dosing schedule of nonpsychotropic medications
OD	12	26	0.58 (0.74)
BD	04	07
TID	02	02
Medications supplied from hospital
No	69 (95.8)	185 (92.5)	0.48 (0.48)@
Yes	3 (4.2)	15 (7.5)
Cost of medicines reimbursed?
Yes	64 (88.9)	186 (93.0)	1.20 (0.27)
No	8 (11.1)	14 (7.0)
Cost of treatment borne by patient/family per month in rupees	580 (324.10)	567 (302.26)	0.31 (0.75)
Does anybody in the family supervise medications
Yes	24 (33.3)	97 (48.5)	4.931 (0.02)*
No	48 (66.7)	103 (51.5)
Type of supervision
Just reminding	12 (50)	53 (57)	0.377 (0.53)
Actual handling over the medicines	12 (50)	40 (43)
Is the patient at time coerced by family members to take medicines?
Yes	11 (15.3)	39 (19.5)	0.629 (0.43)
No	61 (84.7)	161 (80.5)
Nonpharmacological management
Relaxation exercise
Advised	28 (38.9)	115 (57.5)	7.35 (0.007)**
Not advised	44 (61.1)	85 (42.5)
Relaxation exercise
Fully adherent to advice	9 (32.1)	58 (50.4)	20.27 (<0.001)**
Partially adherent to advice	6 (21.4)	45 (39.1)
Nonadherent to advice	13 (46.4)	12 (10.4)
Dietary advice
Advised	13 (18.1)	59 (29.5)	3.56 (0.059)
Not advised	59 (81.9)	141 (70.5)
Dietary advice
Fully adherent to advice	1 (7.7)	18 (30.5)	18.62 (<0.001)***
Partially adherent to advice	5 (38.5)	34 (62.7)
Nonadherent to advice	7 (53.8)	7 (6.8)
Behavioral therapy
Advised	5 (6.9)	36 (18.0)	5.05 (0.02)*
Not advised	67 (93.1)	164 (82.0)
Behavioral therapy
Fully adherent to advice	0 (0)	8 (22.2)	4.78 (0.092)
Partially adherent to advice	3 (60)	25 (69.4)
Nonadherent to advice	2 (40)	3 (8.3)
Treatment from any other department of the institute
Yes	5 (6.9)	10 (5.0)	0.38 (0.535)
No	67 (93.1)	190 (95.0)
Any treatment sought from faith healers
Yes	11 (15.3)	21 (10.5)	1.16 (0.281)
No	61 (84.7)	179 (89.5)
Patient/family follow particular religious cult and help sought
Yes	8 (11.1)	20 (10.0)	0.07 (0.790)
No	64 (88.9)	180 (90.0)
Follow-up details
Came to the OPD (with or without caregiver) during the first visit
Alone	11 (15.3)	18 (9.0)	2.19 (0.139)
Came with attendants	61 (84.7)	182 (91.0)
Referred from
Self	25 (34.7)	66 (33.0)	0.07 (0.791)
Relative/physician	47 (65.3)	134 (67.0)
Total duration of treatment contact in days	18.40 (25.70)	248.5 (71.14)	2628 (<0.001)***
Total number of visits during the entire contact period	2.01 (1.02)	7.08 (1.89)	21.68 (<0.001)***
Total number of scheduled follow-up visits	0.72 (0.85)	4.32 (2.00)	495 (<0.001)***
Total number of unscheduled follow-up visits	0.29 (0.54)	1.77 (1.11)	1749 (<0.001)***
Detail workup done
Yes	22 (30.5)	172 (86)	9.67 (0.002)**
No	50 (69.5)	28 (14)
Last follow-up
Junior resident	22 (30.5)	172 (86)	9.67 (0.002)**
Senior resident/consultant	50 (69.5)	28 (14)

SD – Standard deviation; OPD – Outpatient department; @Chi square test with Yate’s Correction

Level of functioning, Attitude toward Medication, Insight, and Medication Adherence of the Dropout Group and Regular Attendee Group

There was no significant difference in the mean GAF score between the two groups. In terms of current level of functioning, compared to “regular attendees” significantly higher proportion of patients in the 'dropout' group had CGI-S rated as markedly to severely ill. The mean DAI-10 score was also significantly higher, indicating better attitude toward medication for the “regular attendee” group. When the attitude toward medication was assessed using SRAQ, those in the “dropout” group had significantly lower average negative score and total score, reflecting lesser positive attitude toward psychotropic medications. Separate ratings were done for the current level of insight and insight at the beginning of the treatment. At the time of assessment, compared to patients in the “regular attendee” group, patients in “dropout” group had poor insight in terms of awareness about achieved effects of medications and overall current level of insight. In terms of past insight (i.e., at the time of starting of treatment), patients of the “dropout” group had significantly poorer insight in the domains of awareness of the achieved effects of medications, social consequences of illness, and overall insight [Table 4].

Table 4

Comparison of level of functioning and severity of illness of the dropout group and regular attendee group

Variables	Frequency (%)/mean (SD)		χ2/t (P)
	Dropout group (n=72)	Regular attendee group (n=200)
GAF Score (current level of functioning)	61.70 (11.27)	60.98 (14.58)	0.382 (0.703)
CGI-S (current severity of functioning)
Normal-moderately ill	8 (11.1)	51 (25.5%)	6.45 (0.01)*
Markedly-severely ill	64 (88.9%)	149 (74.5%)
Mean CGI severity score	3.70 (0.84)	3.90 (1.02)	0.32 (0.70)
DAI total score	0.83 (3.89)	3.45 (0.40)	4370# (<0.001)***
Self-report attitude questionnaire scores
Average positive score	11.02 (3.76)	10.79 (3.39)	4.85 (0.62)
Average negative score	11.20 (4.49)	13.08 (3.50)	3.60 (<0.001)***
Average total score	22.23 (6.02)	23.88 (5.19)	2.20 (0.028)*
Current rating on SUMD Scale
Awareness of mental disorder	2.05 (1.46)	2.03 (1.45)	0.13 (0.89)
Awareness of the achieved effects of medication	2.77 (1.59)	1.95 (1.30)	4.34 (<0.001)***
Awareness of social consequences	2.66 (1.50)	2.55 (1.64)	0.52 (0.59)
SUMD total score	7.50 (3.50)	6.53 (3.57)	1.98 (0.048)*
Past rating on SUMD scale
Awareness of mental disorder	3.88 (1.37)	3.26 (1.69)	2.80 (0.005)
Awareness of the achieved effects of medication	4.13 (1.20)	3.07 (1.47)	5.48 (<0.001)***
Awareness of social consequences	4.27 (1.31)	3.46 (1.59)	3.87 (<0.001)***
SUMD total score	12.30 (2.75)	9.80 (3.78)	5.14 (<0.001)***

#Mann-Whitney U-value. For the “dropout” group, the patients were asked to respond to BARS, considering their medication compliance of the period prior to dropping out of the treatment. In terms of percentage of medications taken, as assessed using BARS, patients in the “dropout” group took significantly less proportion of medications (42.9 [39.9] vs. 92.6 [13.9]; t-test: 15.25 P<0.001***). GAF – Global assessment of functioning; CGI – Clinical Global Impression-Severity of Illness; BARS – Brief Adherence Rating Scale; SD – Standard deviation; SUMD – Scale to assess unawareness of mental disorder

For the “dropout” group, the patients were asked to respond to BARS, considering their medication compliance of the period prior to dropping out of the treatment. In terms of percentage of medications taken, as assessed using BARS, patients in the “dropout” group took significantly less proportion of medications (42.9 [39.9] vs. 92.6 [13.9]; t-test: 15.25 P < 0.001***).

Level of satisfaction and therapeutic alliance

Higher proportion of patients in the “regular attendee” group were fully satisfied with various aspects of treatment process and had better therapeutic alliance, when compared to the patients in the 'dropout' group [Table 5].

Table 5

Comparison of level of satisfaction of the of the dropout group and regular attendee group

Variables	Frequency (%)/mean (SD)		χ2/t (P)
	Dropout group (n=72)	Regular attendee group (n=200)
Where you satisfied with the place and the time at which you were called for the evaluation?
Fully satisfied	47 (65.3)	175 (87.5)	14.7 (<0.001)***
Partial satisfied/unsatisfied or fully unsatisfied	23 (34.7)	25 (13.5)
Where you satisfied with the time given to you to explain your problems to the doctors and other staff?
Fully satisfied	46 (63.9)	154 (77.0)	3.43 (0.06)
Partial satisfied/unsatisfied or fully unsatisfied	24 (36.1)	46 (23)
How much you are satisfied with the ability of the doctor who saw you?
Fully satisfied	39 (54.2)	153 (76.5)	10.9 (0.001)***
Partial Satisfied/unsatisfied or fully unsatisfied	31 (45.8)	47 (23.5)
Taking your whole treatment process into account how much you are satisfied with the services provided by the department of psychiatry?
Fully satisfied	43 (59.7)	152 (76.0)	5.48 (0.02)*
Partial satisfied/unsatisfied or fully unsatisfied	27 (40.3)	48 (24)
Satisfaction with treatment questionnaire total score	6.09 (2.6)	4.92 (1.4)	4.715 (<0.001)***
STAR scores
Positive collaboration	17.60 (4.60)	20.06 (2.67)	4.77 (<0.001)***
Positive clinician inputs	8.59 (2.21)	9.11 (1.91)	1.88 (0.06)
Nonsupportive clinician inputs	2.77 (3.04)	2.87 (2.58)	6565 (0.26)
Mean total score on the scale to assess the therapeutic relationship	29.27 (6.32)	32.04 (4.46)	4.77 (<0.001)***

SD – Standard deviation; STAR – Scale to assess the therapeutic relationship

Treatment after dropping out and the reasons of dropout in the dropout group

During the follow-up assessment, only participants patients (12.5%) reported seeking further treatment for their ailments. Of those who sought took further treatment after dropping out from our centre, took treatment either from a psychiatrist or a physician. When asked for the reason for not continuing care from the psychiatric outpatient, the most common reason reported was remission of symptoms followed by lack of satisfaction and side effects of medication [Table 6].

Table 6

Treatment sought after dropping out and reasons for dropout

Variables	Dropout group (n=72), frequency (%)/mean (SD)
Any treatment received after dropout
Yes	9 (12.5)
No	63 (87.5)
Type of treatment received after dropout
Treatment from psychiatrists/physician	8 (11.1)
From faith healers or traditional practitioners	1 (1.4)
Reasons for the dropout
Distance	4 (5.6)
Remission of symptoms	24 (33.3)
Time factor	6 (8.3)
Private treatment	3 (4.2)
Poor response with treatment	2 (2.8)
Pregnancy and death of family member	1 (1.4)
Patients own negligence	4 (5.6)
Lack of satisfaction of treatment and side effects of drugs	10 (13.9)
Medical illness	3 (4.2)
Poor socioeconomic status	1 (1.4)
Migration to other place	3 (4.2)
Treatment from faith healer	1 (1.4)
Poor family/social support	4 (5.6)
Poor insight into illness	4 (5.6)
Communication gap	1 (1.4)
Side effects of medications	10 (13.9)
Financial problems	2 (2.8)
Death of relatives	1 (1.4)
Fear of psychiatric treatment	1 (1.4)
Fear of ECT and admission	1 (1.4)
Unable to get leaves from job	1 (1.4)

SD – Standard deviation, ECT – Electroconvulsive therapy

DISCUSSION

Outpatient adherence is one of the most important determinants of the outcome of psychiatric disorder. Better understanding and identification of determinants of nonadherence may suggest ways to improve adherence and may help prevent adverse consequences of nonadherence. Further, persons at greater risk of dropping out may be prospectively identified and targeted for measures to improve compliance.

Patients of both the groups did not differ on most of the sociodemographic variables. However, those who dropped out from treatment were older and were more likely to be married. These findings are somewhat contradictory to the findings from the West, which suggest that those who dropped out from treatment are generally younger and are living alone.[3] This possibly suggests sociocultural differences in the treatment adherence. However, the difference in the marital status must be viewed with caution, because in India, marriage is considered to be more universal phenomenon compared to the Western population. Further, those who dropped out were older. Hence, it is quite possible that the difference in marriage could be due to age per se. Previous studies from India have suggested that those who dropped out from treatment are more frequently male,[36] have higher income,[37] belong to rural background, have low income, are illiterate,[39] and reside at a distance >50 km from the hospital.[3637] However, no such differences emerged in this study. This could possibly be due to changing sociodemographics of the country over the years.[4243]

The mean age of onset for the 'dropout' group was significantly higher than the 'regular attendee' group. Studies from the West and previous studies from India have not commented on the age of onset as a variable associated with dropout from treatment. This finding suggests that clinicians should pay more attention to patients with illness of later age of onset for possibility of dropping out from treatment.

The mean duration of illness at the time of the first assessment for the “dropout” group was significantly higher than the “regular attendee” group. This finding is supported by previous studies from India, which also have reported longer duration of illness to be associated with treatment dropout.[3637]

Previous studies from the West have also reported relationship of dropout status with neurotic disorders.[57] Some of the studies from India also have reported fewer patients with psychotic disorders to be in the dropout group.[36] This finding possibly suggests that higher severity of illness may be responsible for continuation of treatment. Other possible reason could be that in patients with psychotic illnesses, the decision-making about treatment adherence more often lies with the caregivers than the patients per se, whereas for neurotic disorders, treatment adherence is more often under the patient's will. Accordingly, future studies must attempt to evaluate the caregiver's contribution to the treatment adherence, who plays an important role in management of patients with severe mental disorders. No significant difference was seen between the two groups on the variables of past psychiatric history, family history of mental illness, prevalence of comorbid physical illness, and history of hospitalization. Previous studies[1617181936373944] in general have also not reported any influence of these variables on treatment dropout. As reported earlier too, the findings of the present study also suggest that higher number of previous hospitalization is associated with treatment dropout.[1294957] This association could be due to other factors such as experiencing more self-stigma, when admitted to a psychiatric inpatient setting. Hence, there is a need to evaluate these associations further.

In the present study, both the study groups did not differ significantly on the number of medications prescribed (both psychotropic medications and total number of medications), dosing schedule, cost of medication, cost of medication being reimbursed, and medication supply from the hospital. Some of these findings are not in keeping with some of the Western data which have reported not having health insurance (where health care is not free at point of delivery)[5] to be associated with treatment dropout. However, the two study groups differed significantly in terms of supervision of medication, with significantly higher percentage of patients in the regular follow-up group being supervised for their medication intake. Studies which have evaluated medication adherence/compliance also suggest that supervision of medications is associated with better medication compliance.[5859] From this finding, it can be hypothesized that there is a bidirectional relationship between adherence to treatment appointments and medication adherence and adherence to one increases the probability of the other. In the present study, it was seen that recommendation of relaxation exercises, dietary measures, and behavior measures was associated with “regular attendees” status. We could not find any existing literature which has looked into this aspect of the initial consultation. This suggests that initial advice should be comprehensive and the clinicians should briefly discuss all the possible treatment modalities and recommend all the strategies which are applicable to a particular patient.

Studies from various part of the world also suggest that highest proportion of patient's dropout during the initial phase of treatment, as was seen in the present study too. The typical dropout pattern is also very similar to the findings reported in previous studies from India.[363941424546] These findings suggest that, if the patients have be retained in the treatment net, it is important to fulfill the needs of the patients and the caregivers during the initial few visits itself; otherwise, there is high chance of treatment dropout.

During the follow-up assessment, only few patients (12.5%) of the dropout group reported seeking further treatment for their ailments. Of those who sought further treatment after dropping out, took treatment either from a psychiatrist or a physician. When asked for the reason for not continuing care from the psychiatric outpatient, the most common reason reported was remission of symptoms and this was followed by lack of satisfaction and side effects of medication. These findings are similar to the existing literature which also suggests higher rate of medication side effects to be associated with poor treatment adherence.[4243] Studies also suggest the presence of transient symptoms[59] and poor treatment satisfaction to be associated with poor treatment adherence.[15]

In the present study, although the two study groups did not differ on the level of functioning scores, compared to the patients in the “regular attendee” group, significantly higher proportion of patients in the 'dropout' group had CGI-S rated as markedly-severely ill. This finding is similar to the previous reports which suggest that those who miss appointment are more unwell and more functionally impaired than those who attend.[5] Previous studies from India too suggest that nearly half of the patients who dropped out either remain in the same clinical state or have worsening of symptoms.[39] These findings suggest that efforts should be made to contact the patients who dropped out from treatment, to improve their outcome.

Findings of the present study, clearly demonstrates, that those who dropped out from treatment, have more negative attitude toward psychotropic medications. One previous study from India suggested that patients those who dropped out have higher rates of medication nonacceptance, apprehension about bad effects of psychotropics and fear of becoming dependent on medications.[39]

In the present study, in term of current level of insight, compared to the patients belonging to the “regular attendee” group, patients belonging to the “dropout” group had poor insight, in terms of “awareness about achieved effects of medications” and “overall current level of insight.” In terms insight in the past (i.e., at the time of initial evaluation), patients of the “dropout” group had significantly poorer insight in the domains of “awareness of the achieved effects of medications,” “social consequences of illness” and “overall insight.” Studies from the West have also shown that poor insight is a consistent predictor of missed appointments/dropout from treatment.[622] These findings suggest that while evaluating new patients, clinicians should give due importance to evaluation of insight and those with poor insight, should receive proper psychoeducation to reduce the chances of dropout from treatment.

In this study, patients in the dropout group had lower medication adherence compared to the regular attendee group. Previous studies also suggest that nonadherence to appointments is also associated with poor medication compliance.[5] This finding suggests that the clinicians should devote enough time during the initial visit to inform the patients about the need for medication compliance.

Findings of this study suggest that patients in the “dropout” group had lower level of satisfaction with the various aspects of treatment process. Studies from the West suggest that missed appointments and treatment dropout is more frequently associated with noncollaborative decision-making, not having an established outpatient clinician and when seen by an inexperienced clinician.[5] Previous studies from India too suggest that long waiting time and attitude of the attending doctors and staff determine the further compliance with appointments.[3944] Accordingly, there is a need to improve the satisfaction level of the patients and the caregivers.

The findings of the present study suggest that the “regular attendees” perceived significantly higher positive collaboration with the therapist and had significantly better overall therapeutic relationship compared to those in the “dropout” group. This finding supports the existing literature which also suggests a close relationship between the quality of therapeutic alliance and treatment adherence.

The present study has certain limitations, and the findings of the present study must be interpreted in light of the same. The limitations of the present study include small sample size and inclusion of diagnostically heterogeneous group. Although this study included patients who dropped out from treatment, ratings obtained for this group were retrospective in nature and could be biased. The study was limited to a group of patient who had telephonic access. It is quite possible that these inclusion criteria could have led to exclusion of most disadvantage group. Future studies must attempt to overcome the limitations of this study.

CONCLUSION

This study suggests that those who dropped out from treatment are significantly older, more likely to be married, have higher age of onset, have longer duration of illness, receive less supervision for medication at home, higher proportion of them continue to remain symptomatic, have more negative attitude toward medications, have poorer insight, have poor medication compliance, are less satisfied with the treatment provided, and have poor quality of therapeutic alliance. Further, it was seen that those who dropped out usually do so during the initial phase of treatment. Hence, to improve appointment adherence, it is important to address the issues of negative attitude toward medications, improve satisfaction with the treatment contact, and enhance therapeutic alliance.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.